Dear Researchers
I want to identify which export category is more important for economic development of the Pakistan. However, i applied the ARDL-ECM model to obtain the long-term elasticities of gpd with respect to various export categories.
But i also want to investigate the long term causality between the exports categories and gdp. However, i have estimated four variables ecm model. And obtain the ECT term but i think it show the causality from all four variables including labore, capital, export and imports but not export only.
However, i am interested to obtain the long term causality for export only. I tried the two variables model including exports and gdp only but did not find the co-integration relationship between these variables but the both varibales are intergrated order 1 means I(1). Please guide me what i do more to obtain the long-term causality
regards
Irshad
ARDL gives you long run CO-INETGRATION NOT CAUSALITY.
CLEAR YOUR POINT
Dear Farrukh Mahmood
In the t-statistic value of the restricted ECM-ARDL model show the long-term causality. A huge literature have used this method for causality analysis . What you suggest which method i will used to obtain the long-term causality
regards
Dear Sayed Irshad Hussain
I agree with your research goals, Sayed. However, one must acknowledge that there are fundamental problems with the idea of isolated causality (the long-term causality associated with only one variable). The true is that the economist's world is made with multiple-causality laws (please take a look to the John Hicks' classic on "Causality in Economics").Therefore, I would advise to slightly modify the goal.Furthermore, under the view of production chains, it is quite posible that there may be a 2-sector or a 3-sector export category network which are better related to gdp (economic development) than any individually consider sector. This is another important way to modify the research goal (and to make it more complex and surely more relevant). All the best, Carlos Barrera
Dear Carlos
I mean from individual the causality between exports and gdp. Most of the literature done in bi variate framework. but they obtained the causality in short-run but same i want in long-run. please comments
Dear Sayed
I hope I do not sound too pretentious while evaluating the literature. Clearly, bivariate relationships may provide some insights, but I must say that bigger models (either VAR or VEC, with some additional work seeking parsimony a' la Luthkepohl) will provide better insights and unfortunately, we do not see those types of articles being published. This a huge problem in the economics-econometrics literatur. Those papers you mentioned should acknowledge the limits of the parametric models they use while searching for empirical laws. I repeat: we must deal with multivariate causality (be it of long-run type or not).
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Climate Science
- Climate Change